Electronic Conductivity of Semiconductor Nanoparticle Monolayers at

May 21, 2003 - The electronic conductivity of PbS and CdTe nanoparticle ... Citation data is made available by participants in Crossref's Cited-by Lin...
0 downloads 0 Views 129KB Size
J. Phys. Chem. B 2003, 107, 5733-5739

5733

Electronic Conductivity of Semiconductor Nanoparticle Monolayers at the Air|Water Interface Ivan A. Greene,† Fanxin Wu,‡ Jin Z. Zhang,‡ and Shaowei Chen*,† Department of Chemistry and Biochemistry, Southern Illinois UniVersity, Carbondale, Illinois 62901, and Department of Chemistry and Biochemistry, UniVersity of California, Santa Cruz, California 95064 ReceiVed: December 11, 2002

The electronic conductivity of PbS and CdTe nanoparticle monolayers was examined voltammetrically by using interdigitated array (IDA) electrodes at the air|water interface. Their band gap energies were estimated from the I-V responses and were very consistent with results obtained from optical measurements as well as solution electrochemistry. For CdTe nanoparticles, the I-V responses were analogous to those of a molecular diode with reproducible voltammetric behavior after repeated potential cycling. Interestingly, there appeared to be indications of particle surface trap states in the voltammetric responses that correlated with spectroscopic measurements. In addition, the band gap of the nanoparticle monolayers could be manipulated by the interparticle interactions, red shifting with decreasing interparticle separation. In contrast, the electroactive nature of the PbS particles led to the decomposition of the nanoparticles and hence deposition onto the electrode surface. The resulting voltammetric responses evolved from those typical of the faradaic reactions to a rectifying feature of much larger current scales, which finally became linear (ohmic) because of shorting between neighboring IDA fingers. In these studies, it was found that photoexcitation played an important role in regulating the current responses, providing a mechanistic basis on which to manipulate the electronic/electrical properties of semiconductor nanomaterials. The conductivity of the final interfinger deposits was about 2 orders of magnitude smaller than that for pure metallic lead, indicating some surface contamination and/or less than perfect crystalline structure.

Introduction The recent intense interest in nanoscale materials is mainly driven by their unique properties that can be easily manipulated by their physical dimensions and surface morphology as well as their chemical environment.1 In addition, in organized ensemble structures, the distribution and ordering of the particle molecules play an important role in regulating the electronic properties of the overall assemblies.2 These will be the key structural parameters in the fabrication of novel electronic nanodevices and nanocircuits. For instance, Heath and coworkers measured the nonlinear optical properties of a Langmuir monolayer of alkanethiolate-protected silver nanoparticles and observed an insulator-metal transition when the interparticle spacing was reduced by mechanical compression.3 This was interpreted by the distance dependence of electronic coupling between neighboring particle molecules. More recently, using an interdigitated array (IDA) electrode, we directly measured the electronic conductivity of alkanethiolate-protected gold nanoparticles at the air|water interface.4 For particles with protecting monolayers of short-chain ligands, we observed ohmic current-potential (I-V) responses whereas for longer chain lengths nonlinear I-V curves were generally found to have rectifying character. This discrepancy in electronic properties could not be interpreted solely on the basis of interparticle distance. More likely, it was related to the combined effects of organic insulating layers on particle electronic interactions as * To whom all correspondence should be addressed. E-mail: schen@ chem.siu.edu. † Southern Illinois University. ‡ University of California.

well as on a variation in electron-transfer mechanisms (e.g., tunneling, hopping, or metallic). In addition, these observations are in sharp contrast with previous studies of solid-state conductivity measurements using drop-cast (µm) thick films of nanoparticles where generally only ohmic responses were found.5,6 This latter observation was, at least in part, attributable to the structural inhomogeneity in the particle thick films where effective electron-transfer pathways might be facilitated by film defects. Previous efforts have been mainly focused on transition-metal nanoparticles, whereas studies of the electronic conductivity of semiconductor nanoparticle materials are relatively scarce.7,8 For instance, Otten et al.7a measured the current noise spectra of PbS nanoparticle thin films, which indicated a random walk (diffusion) of electrons between the particles. Alperson et al.7b used a conductive scanning force microscope to investigate the electronic conductance of isolated CdSe quantum dots where the nanoparticle band gap as well as Coulomb charging were evaluated. Mallouk and co-workers8 fabricated a nanoscale heterojunction with semiconductor nanoparticles and observed rectifying current responses. Of these studies, one of the intriguing properties associated with semiconductor nanoparticles is their band gap energy, which has been found to be sensitive to particle dimensions because of quantum confinement effects as well as to the chemical environment of the particles. Because a variety of nanoparticle properties (optical, luminescence, electronic, etc.) are dependent upon this band gap energy, an accurate assessment is of paramount importance in understanding the molecular mechanism. Typically, this energy is characterized by optical measurements (e.g., UV-vis spectros-

10.1021/jp027692t CCC: $25.00 © 2003 American Chemical Society Published on Web 05/21/2003

5734 J. Phys. Chem. B, Vol. 107, No. 24, 2003 SCHEME 1

Greene et al.

a

a (A) Experimental setup for electronic conductivity measurements at the air|water interface (not to scale).4 (B) Schematic of the IDA fingers and the nanoparticle monolayers.

copy).9 Recently, it was found that electrochemistry was also an effective complementary tool where the nanoparticle band gap was reflected by a featureless current profile in voltammetric measurements.10,11 However, for many nanosized semiconductor materials, the band gap typically lies in the range of a few electron volts, sometimes rendering it difficult to evaluate this energy structure in solutions using conventional electrochemical approaches because of limited access to suitable solvents. Solidstate electrochemical approaches offer an effective alternative. In this article, we will report on the electronic conductivity measurements of semiconductor nanoparticle monolayers at the air|water interface by using monolayer-protected PbS and CdTe nanoparticles as the illustrating examples. We will first focus on their solid-state electron-transfer chemistry and then examine the effects of interparticle separation on the band gap energies of nanoparticle ensembles. More significantly, we will demonstrate that nanoparticle trap states can be located by combining voltammetric results with spectroscopic data. Experimental Section The synthesis of n-hexanethiolate-protected PbS nanoparticles has been described in detail in a previous report.11 CdTe nanoparticles used in the present study were stabilized first by a monolayer of thioglycolic acid in aqueous solutions and rendered hydrophobic by binding to a second layer of dimethyldioctadecylammonium. The synthesis of thiol-capped CdTe in aqueous solutions was also detailed in earlier articles.12 The average core sizes of these two nanoparticles were ca. 4 and 2 nm, respectively, with a very narrow dispersity (standard deviation of about 15% of the average particle size), as determined by transmission electron microscopy. UV-vis absorption spectra were acquired with a Unicam ATI UV4 spectrometer, and fluorescence studies were carried out with a PTI fluorescence spectrometer. The particle solutions were prepared in CHCl3 at a concentration of approximately 2.75 µM. In both cases, a monolayer of the nanoparticle molecules was formed at the air|water interface using the Langmuir technique (NIMA 611D). For PbS nanoparticles, typically 250 µL of the particle solutions (2.75 µM CH2Cl2) was spread dropwise onto the water surface. At least 20 min was allowed for solvent evaporation as well as between compression cycles. An interdigitated array (IDA, from ABTECH Scientific) electrode was aligned vertically at the air|water interface where a monolayer

Figure 1. Langmuir isotherm of PbS nanoparticles. n-Hexanethiolateprotected PbS nanoparticles were dissolved in CH2Cl2 at a concentration of 2.75 µM. This solution (250 µL) was spread dropwise onto the water surface. Compression speed 20 cm2/min.

of nanoparticles was trapped between IDA fingers. (Details of the experimental setup are shown in Scheme 1, as described previously4). The IDA electrode consists of 25 pairs of gold fingers with dimensions of 3 mm × 5 µm × 5 µm (L × W × H). The corresponding current-potential (I-V) profiles were collected directly at the air|water interface by applying a voltage bias to the contacts of the finger pairs using an EG&G PARC potentiostat (model 283) and EG&G commercial software (PowerCV). Results and Discussion Figure 1 shows the Langmuir isotherm of PbS nanoparticles on the water surface. One can see that at trough areas greater than 26 nm2/particle the surface pressure is essentially zero, equivalently indicating a 2D gaseous state of the particle molecules whereas at a smaller surface area the pressure starts to rise rather rapidly, suggesting that the particles are in close contact and ligand intercalation starts to occur. At this takeoff area, on the basis of a hexagonal arrangement within the particle monolayers, the average interparticle center-to-center distance can be estimated to be about 5.61 nm, which is only slightly larger than the physical diameter of the PbS particles (core +

Conductivity of Nanoparticle Monolayers

J. Phys. Chem. B, Vol. 107, No. 24, 2003 5735

Figure 2. (A) Current-voltage profiles of PbS nanoparticle monolayers at π ) 3 mN/m in the dark. Potential scan rate 10 mV/s. (B) Effects of photoexcitation (red light) on the electronic conductivity of PbS nanoparticles. Potential scan rate 20 mV/s. (C) Current-potential profiles in the dark after photoexcitation as described in B. Potential scan rate 20 mV/s. The inset in A depicts the potential program, where E+,lim and E-,lim denote the positive and negative limits of the potential range, respectively.

two fully extended chains of hexanethiolates with 0.78 nm each as calculated by Hyperchem). The overall behaviors are quite similar to those of alkanethiolate-protected gold nanoparticles.4 The corresponding current-voltage profiles are shown in Figure 2A with varied potential windows (with surface pressurecontrolled at 3 mN/m, i.e., the interparticle edge-to-edge spacing (L) is about 0.97 nm). The potential scanning program is shown in the Figure 2A inset. One can see that within the potential range of -1.0 and +0.8 V the current response (s), on the order of a few nanoamps, is essentially featureless. However, when the negative potential is expanded to -1.6 V (s), a cathodic peak appears at around -1.45 V with no apparent return wave. In addition, when the positive potential is simultaneously expanded to +1.0 V (. . .), an anodic peak starts to show up at ca. +0.7 V, and on the return (cathodic) scan, a new peak appears at -0.5 V in addition to the original cathodic peak observed in the previous potential window (which now shifts to -1.25 V). These behaviors are very similar to those observed in electrochemical studies of PbS nanoparticles in solutions.11 These voltammetric features are ascribed to the charge-transfer reactions involved in PbS decomposition. It has been rather well known that PbS undergoes decomposition

processes at very negative and positive potentials by the following reaction mechanisms:11

anodic dissolution cathodic reduction

PbS f Pb2+ + S + 2e +

PbS + 2H + 2e f Pb + H2S

(1) (2)

Thus, one can see that the voltammetric peak at -1.25 V is most likely related to the cathodic reduction of PbS nanoparticles (2) leading to the deposition of metallic Pb onto the electrode surface whereas the anodic peak at +0.7 V is attributable to the anodic dissolution of PbS nanoparticles (1) and the voltammetric peak (-0.5 V) on the subsequent cathodic scan might be ascribed to the formation of Pb from Pb2+ generated in reaction 1. In contrast, when the electronic conductivity was measured in vacuo on a solid substrate, PbS nanoparticles were found to be stable even at a much higher voltage bias (e.g., up to 10 V).7a It should be noted that the above interpretation is oversimplified because the alkanethiolate ligands have not been taken into account. Because of the hydrophobic nature of these alkanethiolate molecules, it is most likely that they stay at the air|water interface, and some might be even adsorbed to the IDA gold finger surfaces.

5736 J. Phys. Chem. B, Vol. 107, No. 24, 2003 From these I-V measurements, one can also estimate the PbS nanoparticle band gap energy. For instance, Figure 2A depicts a flat current profile within the central potential region from -1.0 to +0.8 V, indicating a band gap of about 1.8 eV. This is very consistent with that evaluated voltammetrically in PbS nanoparticle solutions.11 In addition, the roughly symmetric current onset about the zero voltage bias suggests that the Fermi level is located in the middle of the gap (at zero voltage bias).7 More interestingly, photoexcitation appears to exert rather substantial impacts on the electronic conductivity of the PbS nanoparticle monolayers. Figure 2B shows the variation of the particle voltammetric currents (within the potential range of -1.6 to +1.0 V) with and without exposure to a red light source (650 nm,